Health Articles

 

C: documents and settings eohcnm.001 local settings temporary internet files content.ie5 2ljs187y s59[1].pdf

⏐ PUBLIC HEALTH MATTERS ⏐ What Has a Decade of Daubert Wrought? Margaret A. Berger, JD Demands for tort reform may also have There have been changes within the judicial system that may be attributable to played a part in the Court's willingness to ac- opinions on the admissibility of expert testimony that began with the Supreme cept the Daubert case for review. There had Court's 1993 decision in Daubert v Merrell Dow Pharmaceuticals, Inc. After sur-veying Daubert and subsequent related Supreme Court opinions, I examine a been an enormous increase in product liabil- number of questions. ity cases and toxic tort litigation in the previ- Do the factors courts apply post-Daubert in ruling on the admissibility of expert ous 15 years or so. Asbestos litigation was testimony make scientific sense? Has Daubert had an impact on the willingness already threatening bankruptcies and unman- of scientists to become expert witnesses? What do we know about Daubert's ageable court congestion. By the time the Su- impact on improving science in the court room? What has been Daubert's effect preme Court undertook to hear Daubert, on access to the courts? Does Daubert further public policy objectives of pro- plaintiffs' experts were being castigated with tecting the public against harm? (Am J Public Health. 2005;95:S59–S65.
some frequency as the villains whose testi- mony, supposedly based on "junk science,"3was responsible for huge unjustified verdicts THE SUPREME COURT TRILOGY ON involve constitutional issues, the interpreta- in product liability and toxic tort actions. The tion of federal statutes, or the review of ad- phrase in quotes, coined in a book published ministrative decisions, which are the kinds of in 1991, quickly became a shorthand expres- Ten years ago, the U.S. Supreme Court de- cases that lie at the heart of Supreme Court sion for referring to perceived problems with cided Daubert v Merrell Dow Pharmaceuticals, jurisprudence. That the Supreme Court is- expert witnesses.
Inc,1 the first in a trilogy of cases about sued three opinions on expert testimony in Given this climate, it is probably not sur- the admissibility of expert testimony. [The civil cases over a 6-year period obviously prising that Daubert was a toxic tort case, as other two cases are General Electric Co v suggests a great deal of interest in this topic.
was General Electric v Joiner, the second case Joiner, 522 US 136 (1997) and Kumho Tire In addition, the Supreme Court decided a in the trilogy. Toxic tort cases are a subspecies Co v Carmichael, 526 US 137 (1999).] To un- fourth case, Weisgram v Marley Co, 528 US of product liability litigation in which the derstand the significance of this case, which 440 (2000), which has an indirect effect on plaintiffs claim that their adverse health ef- some deem one of the more important deci- expert testimony that is discussed infra at fects were caused by exposure to the defen- sions of the twentieth century,2 we must first notes 29–30.
dant's product. The third case in the trilogy, appreciate the central role of expert witnesses We can, of course, only conjecture about Kumho Tire Co v Carmichael, though not a in civil litigation. As the modern world be- the impetus for the Court's foray into this toxic tort case, was a product liability case.
comes increasingly complex, more and more area of the law. Certainly, the growing depen- The crucial issue in all these cases was causa- issues in legal disputes cannot be resolved dence on technology and science in our soci- tion. To prevail, plaintiffs in each case, through without recourse to specialized knowledge.
ety meant that more issues turning on expert the offer of expert testimony, had to dis- The necessary input is supplied through the testimony were entering the courtroom. For charge their burden of proving that the defen- testimony of expert witnesses. Plaintiffs bear example, the explosive growth of computers dant's product had caused the plaintiffs' in- the burden of proving their claims are more required parties to present complex forms of juries. Because plaintiffs' expert testimony probable than not; they cannot succeed when data and statistical studies needing expert elu- was excluded, plaintiffs lost.
expertise is required to support their claims cidation. In addition, the revolutionary advent Daubert was one of a series of cases in unless they are allowed to introduce expert of forensic DNA technology, which was first which plaintiffs claimed that their serious testimony. By setting out new rules for when introduced in an American courtroom just a birth defects stemmed from their mothers an expert may testify, Daubert has therefore few years before Daubert, undoubtedly drew having taken a drug called Bendectin, an had an enormous effect on the outcome of the Supreme Court's attention to how science anti–morning sickness drug that had been and law interact. In addition, numerous pres- taken by over 20 million women. As a result Before we look at Daubert and the other tigious groups, including the Federal Courts of the litigation, the defendant took the drug Supreme Court's opinions in more detail, I'd Study Commission established by Congress, off the market, although it never lost its Food like to speculate for a moment on what led the Judicial Conference, and the Carnegie and Drug Administration approval and con- the Supreme Court to plunge into the realm Commission on Science, Law and Technology, tinued to be available in Canada. In support of expert proof. After all, the Supreme Court had begun actively calling for a reexamina- of its contentions, plaintiffs relied on in vitro reviews only about 80 cases in a given year.
tion of how courts handle complex scientific and in vivo studies, pharmacological studies Furthermore, many of the cases it does hear and technological issues.
of the chemical structure of Bendectin, and Supplement 1, 2005, Vol 95, No. S1 American Journal of Public Health Berger Peer Reviewed Public Health Matters S59 ⏐ PUBLIC HEALTH MATTERS ⏐ their experts' reanalyses of previously pub- tion if it dealt with a matter that was not at clude the plaintiffs' expert. One such case, in lished epidemiological studies. The lower issue. The Court explained its concept of fit which the trial court granted the defendant's courts hearing Daubert relied on the so-called by citing the Third Circuit's opinion in United motion and then dismissed the action, was Frye or "general acceptance" test to hold that States v Downing, 753 F2d 1224,1242 (3d brought by a 37-year-old plaintiff named the plaintiffs' expert testimony on causation Cir 1985). In Downing, the defendant had Joiner against General Electric. Joiner, who was inadmissible and that, consequently, sought to introduce expert testimony about was a heavy smoker with a family history of plaintiffs "could not satisfy their burden of the unreliability of eyewitness testimony. The lung cancer, claimed that exposure to poly- proving causation at trial."4 court held that such evidence was admissible chlorinated biphenyls (PCBs) and their deriv- The Frye test, which gets its name from if it "fit" the facts of the case. On remand, the atives had promoted the development of his Frye v United States, a 1923 federal case in- district court excluded the expert's testimony small-cell lung cancer. The appellate court re- volving the admissibility of polygraph evi- about problems with cross-racial identifica- versed because it found that a very stringent dence,5 was then used by some federal tion because there had been no cross-racial standard of review had to be used when the courts, primarily in criminal cases, and is still exclusion of evidence was outcome determi- in use in some of our most populous states, Perhaps of paramount importance, the native, that is, when the refusal to permit the such as California, Illinois, and New York.
Daubert opinion recast the role of the trial plaintiff's expert to testify would result in dis- Frye conditions the admissibility of expert tes- court. Trial judges had always had the power missal of the case. The Supreme Court, how- timony about a novel scientific principle on to exclude inappropriate expert testimony, ever, held unanimously in Joiner, that regard- there being a consensus, or "general accept- but some preferred to leave this task to the less of whether the trial court admits or ance," of the underlying theory in the rele- jury, particularly when the expert proof re- excludes evidence, the appellate court must vant field. The Frye test has been criticized lated to complex scientific principles with use an abuse of discretion standard in review- on numerous grounds, for example, that it which the judge was not very familiar or com- ing a trial court's evidentiary ruling. This stan- fails to explain how to determine what is the fortable. But the Court now told trial judges dard means that the reviewing court may not relevant field, that it counts the noses of ex- that they were "gatekeepers" who were look at the ruling de novo and decide how it perts rather than looking at the validity obliged to screen scientific expert testimony would have ruled under the circumstances; of their opinions, and that it leads to self- for relevancy and reliability before it could instead, it must defer to the rulings of the trial validating experts who claim that their partic- be admitted.
court unless they are manifestly erroneous.
ular subspecialty is the relevant field.
The Supreme Court did not apply its new Accordingly, in Joiner, the Supreme Court In Daubert, the Supreme Court first found test for the admissibility of expert testimony examined the record to determine whether that the Frye test was superseded when the in the Daubert case. Instead, it reversed the the district court had abused its discretion Federal Rules of Evidence, which govern evi- decision and remanded the case to the lower when it excluded the plaintiff's expert testi- dentiary questions in federal court, were en- court. The day after the Supreme Court de- mony. The Court found that the trial court acted in 1975 and failed to mention Frye. cided Daubert, the Wall Street Journal stated had not erred when it concluded that the Second, the Court set out a new two-pronged that plaintiffs had won because the Ninth Cir- plaintiff's experts had not explained "how and test for the admissibility of scientific evidence, cuit had been reversed for using the wrong why" they could extrapolate proof of causa- whose object was to ensure that expert testi- test 6; the New York Times, which better un- tion from animal studies conducted under cir- mony "is not only relevant, but reliable."5 derstood the implications of the opinion, pre- cumstances that differed from the conditions Justice Blackmun, who wrote the majority dicted that defendants were the true victors.7 surrounding the plaintiff's exposure. The opinion for the Court, explained that in order On remand, the Ninth Circuit again excluded studies involved infant mice injected with to satisfy reliability, the expert must have de- the evidence and granted summary judgment massive doses of PCBs; the plaintiff was ex- rived his or her conclusion by the scientific for the defendant because the plaintiffs were posed as an adult through physical contact method; he noted the following factors, unable to prove causation.8 (Rule 56 of the with fluids containing far lower concentra- which, though not definitive, operate as Federal Rules of Civil Procedure provides that tions of PCBs, and the mice and the plaintiff markers of the scientific method: hypothesis a court may grant summary judgment and developed different forms of cancer.
testing, peer review and publication, known dismiss a case before trial when there is no The Court further found that the trial court or potential rates of error, and the existence genuine issue as to any material fact. When had not abused its discretion when it rejected of standards controlling the technique's oper- plaintiffs cannot prove a material fact such as the plaintiff's epidemiological evidence. Ac- ation. General acceptance of the methodology causation because their experts have been ex- cording to the district court, the authors of in the relevant discipline, although no longer cluded, there is no longer a disputed issue for one study had refused to conclude that PCBs dispositive, was mentioned as a factor also to the jury to resolve.) were the cause of a somewhat higher than ex- be considered. The second prong—relevancy— Of course, Daubert was read by lawyers as pected rate of lung cancer at an Italian plant; meant that the expert's theory had to fit the well as judges. Defense counsel quickly began the results of another study were not statisti- facts of the case. Even if the expert's theory making motions—soon known as "Daubert cally significant; a third study did not mention was completely scientific, it had no applica- motions"—prior to trial asking the court to ex- PCBs; and the workers in a fourth study S60 Public Health Matters Peer Reviewed Berger American Journal of Public Health Supplement 1, 2005, Vol 95, No. S1 ⏐ PUBLIC HEALTH MATTERS ⏐ considered by the trial judge had been ex- studies or personal experience, employs in continuing legal education programs, and a posed to numerous other possible carcino- the courtroom the same level of intellectual vast law review literature. Even in jurisdic- gens. Consequently, the Supreme Court found rigor that characterizes the practice of an tions that purportedly follow the Frye "gen- that the trial court could properly conclude expert in the relevant field." eral acceptance" test, judges are citing and an- that "the studies upon which the experts re- But the Court declined to articulate a rigid alyzing Daubert and its progeny, and many lied were not sufficient, whether individually classification system for different fields of ex- states have adopted Daubert.
or in combination, to support their conclu- pertise. The Court refused to find "a schema- sions that Joiner's exposure to PCBs con- tism that segregates expertise by type while Impact on Toxic Tort Litigation tributed to his cancer."9 mapping certain kinds of questions to certain If we look at toxic tort cases, we see that The third Supreme Court case, Kumho, kinds of experts. Life and the legal cases that numerous problems arise when courts seek dealt with the admissibility of engineering tes- it generates are too complex to warrant so to determine, as commanded by Daubert, timony offered to prove that the blowout of a definitive a match." The Court explained that whether the expert's conclusions were tire on the plaintiff's minivan was caused by a the Daubert factors are not always relevant reached through the scientific method. Some defect that brought about an accident in even when the expert relies on scientific evi- commentators have expressed doubt about which one passenger was killed and others dence. The Court emphasized that "Too the ability of courts to understand scientific suffered serious injuries. The plaintiff's expert much depends on the particular circum- principles and have suggested that some intended to testify, as he had during a deposi- stances of the particular case at issue." Quot- judges simply use the Daubert factors as a tion, that on the basis of a visual inspection ing from the brief for the United States as checklist without understanding how they re- he could tell that the tire had not been Amicus Curiae, the Court explained that ad- late to reliability.11 Some support for these abused and therefore must have been defec- missibility will depend "on the nature of the critiques was furnished by a recent survey of tive. When the defendant made a Daubert issue, the expert's particular expertise, and 400 state court judges that concluded that motion, the trial court originally excluded the the subject of his testimony." the participating judges had little understand- expert on the ground that the four Daubert Although nothing in the Kumho opinion is ing of the key concept of hypothesis testing factors—testability, peer review or publication, inconsistent with Daubert, the Court's opinion or of the significance of error rates, although known error rates and standards, and general does seem to set out a more flexible test. In- they did considerably better with the other acceptance—had not been satisfied. On recon- stead of stressing factors that, although not two Daubert criteria of peer review and gen- sideration, the trial judge conceded that he definitive, are nevertheless suggested as eral acceptance.12 However, critics have also had erred in treating the factors as mandatory guides for determining reliability, Justice questioned the appropriateness of the Daubert rather than illustrative indicators of reliability, Breyer in Kumho stressed the need to look at factors as guides to reliability. A recent issue but he nevertheless again excluded the plain- reliability in the context of the particular case of the Journal of the American Medical Associ- tiff's expert and granted summary judgment.
and the testimony being offered. Courts, how- ation raises serious doubt about the efficacy The intermediate appellate court reversed ever, are citing and relying on Daubert more of peer review in assuring the reliability of on the ground that Daubert applied only to tes- frequently than Kumho. (A Westlaw search on scientific publications.13 As peer review ap- timony that relies on the application of scien- June 28, 2004, found 2708 citations in judi- pears to be a criterion that judges understand, tific theories or principles and not to testimony cial opinions to Daubert since Kumho was de- it may be that they are relying too heavily on based on the expert's "skill- or experience- cided and only 1454 citations to Kumho.) a problematic marker of good science.
based observation."9 It was the split in the In most instances, plaintiffs seeking to es- courts about Daubert 's applicability to nonsci- THE IMPACT OF THE TRILOGY tablish causation in a toxic tort case do not entific evidence that was the Supreme Court's suffer from a "signature disease" that is stated reason for reviewing the Kumho case.
Having examined what the Supreme Court uniquely associated with exposure to the de- All the justices, in an opinion by Justice said in these three opinions, I now turn to the fendant's product. Instead, they have experi- Breyer, agreed that the trial court's gatekeep- trilogy's effects, first in toxic tort cases, and enced adverse health consequences that also ing obligation to screen expert testimony be- then on litigation in general. (This is not to affect those who have not been exposed to the fore it may be admitted extends to all expert suggest that Daubert is not having an impact substance in question and have no explana- testimony. The Court noted that the govern- in numerous other fields. Engineers, econo- tion for how the exposure caused their dis- ing rule of evidence "makes no relevant dis- mists, psychologists, and numerous other ease. (In some instances, plaintiffs have a tinction between ‘scientific' knowledge and types of experts are also being excluded.) causal explanation, as in the Dalkon Shield ‘technical' or ‘other specialized' knowledge" Clearly, Daubert is viewed as extremely im- cases in which plaintiffs claimed that the and "applies its reliability standard to all . .
portant. There are now Daubert databases string on the shield wicked bacteria into the matters within its scope."10 The Court stressed that collect judicial opinions that already wearer's uterus.) As in Joiner, plaintiffs will that the gatekeeping role requires the trial number in the thousands, at least one loose- typically rely on a combination of epidemiol- court to "make certain that an expert, leaf service that deals exclusively with expert ogy and animals studies to prove causation.
whether basing testimony upon professional witness testimony, countless symposiums and Courts have expressed a preference for the Supplement 1, 2005, Vol 95, No. S1 American Journal of Public Health Berger Peer Reviewed Public Health Matters S61 ⏐ PUBLIC HEALTH MATTERS ⏐ epidemiologic study because it relates to Torts, January 21, 2003: [Epidemiologists ex- would look at the totality of the evidence be- human effects. But such studies may not be perience] "inarticulateness in expressing the fore reaching a conclusion.
available when the plaintiff needs to institute levels of uncertainty below classical levels of Parlodel cases. The different approaches an action in order to avoid having the suit surety"; epidemiologists are not familiar courts take on the admissibility of expert barred by the statute of limitations. Or it may using language "to express and navigate this proof in toxic tort cases is exemplified by the be impossible to do an epidemiologic study very, very difficult, sort of intermediate terri- inconsistent results in the line of cases involv- because the product has been taken off the tory, which is exactly the territory where ing Parlodel. Parlodel is a drug that was ap- market or because the condition from which many of these suits are occurring as evidence proved in 1980 to suppress lactation in the plaintiff suffers is so rare. In addition, such is accumulating, as you say, you have very women who did not wish to nurse. In a num- studies are expensive and time-consuming and little epidemiologic evidence.") (Statement of ber of instances, women taking the drug suf- subject to confounders and biases.14 Neverthe- Steven Goodman.) Because the scientific ob- fered strokes or heart attacks within a few less, some courts have suggested that plaintiffs jective is to obtain the most knowledge possi- days of commencing the medication. Reports cannot win in the absence of a positive epi- ble, the system deliberately chooses to err on of these incidents led the FDA to conclude demiological study, even when the defendant the side of wrongly rejecting the hypothesis that Parlodel should not be used to prevent has far more available information and re- being tested—drug A causes disease X— lactation. These occurrences also resulted in sources to undertake such a study, and even rather than erroneously finding that it was law suits in both federal and state courts. The though the scientific community looks to other proven.15 But the fact that a hypothesis was results have been mixed. Some courts have forms of proof in assessing causation, especially not proven does not mean that it was dis- permitted plaintiffs' experts to testify [see, for animal studies. Some courts appear to find all proved. If a study fails to achieve a 0.05 example, Globetti v Sandoz Pharmaceuticals animals studies irrelevant because of the need level of statistical significance—a convention Corp, 111 FSupp2d 1174 (ND Ala 2000)]; to extrapolate to humans and because, to see often used to ensure the stringent scientific but others have excluded the experts and an effect, the doses to which the animals are standard of proof—that does not mean it has granted summary judgment [see, for example, subjected must be many times higher than no probative value. To a scientist, it simply Siharath v Sandoz Pharmaceuticals Corp, 131 those administered to humans. See, for exam- means that more research may be in order.
FSupp2d 1347 (ND Ga 2001), aff'd, Rider v ple, Wade-Greaux v Whitehall Labs, Inc, 874 Some courts, however, read Daubert to mean Sandoz Pharmaceuticals Corp, 295 F3d 1194 FSupp 1441,1480 (DVI 1994), in which the that expert testimony that does not meet this (11th Cir 2002)]. The plaintiffs offered an ex- court stated: "The notion that one can accu- scientific standard is inadmissible, even planatory theory of the causal process—that rately extrapolate from animal data to human though a plaintiff in a civil case only has the the drug, like other ergot alkaloids, can cause to prove causation without supportive epide- burden of proving its case by a preponder- vasoconstriction and hypertension, which in miologic studies is scientifically invalid be- ance of the evidence. That is, the plaintiff's turn can cause seizures and strokes. They cause it is inconsistent with several universally burden is to convince the trier of fact on the also relied on animal studies, adverse reaction accepted and tested scientific principles. The basis of all the evidence that was admitted reports, opinions by clinicians and, probably principle of species specificity has been tested that it is more probable than not that defen- most importantly, the results of a number of and demonstrates that different species can dant did what was alleged in the complaint.
de-challenge/re-challenge experiments in react differently to the same agent." Compare These courts condition admitting an ex- which women given Parlodel developed rele- Joint Discussion of Science, Technology, and pert's testimony on there being evidence vant symptoms that ceased when the drug Law Panel and American Law Institute, Re- sufficient from a scientific standpoint to was discontinued and recommenced when statement of Torts, January 21, 2003: "There prove the hypothesis being tested, which is the drug was reinstituted. The little epidemio- are an extraordinarily few number of places a standard of proof that far exceeds that ap- logical evidence available was inconclusive.
in the literature where anyone has ever even plicable to civil litigation.
The courts that excluded plaintiffs' experts asked the question, if it happens in mice, does Another problem confronting plaintiffs in in the Parlodel litigation simply found that it happen in people? Is an animal model, an toxic tort cases is that courts often do what this evidence did not suffice to prove causa- animal [study] that is done in a mouse or a the district court in Joiner did. It examined tion. They termed the adverse reaction re- rat, how is that comparable to humans?" and rejected separately each study on which ports anecdotal and were unimpressed with (Statement of Douglas Weed.) the plaintiff's expert relied rather than look- the testimony proffered by clinicians. One The standard of proof. Courts also fail to ing jointly at all the scientific studies on which court stated that "with regard to general cau- recognize that in reporting the results of their the plaintiff's experts based their opinions.
sation, the relevant scientific field is epidemi- studies, epidemiologists use a far more strin- The Supreme Court in Joiner seemed to ap- ology or toxicology and not clinical medi- gent standard than the preponderance of the prove this practice because it found that the cine." [See Siharath v Sandoz Pharmaceuticals evidence standard that applies in civil litiga- trial court did not abuse its discretion in find- (2001).] And yet the Supreme Court in tion. See, for example, Joint Discussion of ing the studies insufficient, "whether individu- Kumho proposed as a test "that an expert, Science, Technology, and Law Panel and ally or in combination." But scientists, includ- whether basing testimony upon professional American Law Institute, Restatement of ing epidemiologists, typically state that they studies or personal experience, employs in S62 Public Health Matters Peer Reviewed Berger American Journal of Public Health Supplement 1, 2005, Vol 95, No. S1 ⏐ PUBLIC HEALTH MATTERS ⏐ the courtroom the same level of intellectual note stated that "if the FDA asks for bad news, causation if they sued for their alleged in- rigor that characterizes the practice of an ex- we have to give, but if we don't have it, then juries. Possibly, the future outcome and im- pert in the relevant field." Although it is true we can't give it to them."19 pact of the Desiano case may shed some light that in many cases, a clinician would not be The difficulty plaintiffs face in proving cau- on this question. Desiano is a consumer fraud concerned with what caused a patient's dis- sation may make the corporation's risk of lia- class action suit, not a toxic tort case, that was ease, because treatment would in no way be bility seem quite minimal when it first learns brought by health insurers who allege that affected by the answer to that question, a cli- of adverse reactions and, besides, even if the cost of Rezulin, which was three times nician may have to make a judgment about plaintiffs ultimately succeed in proving causa- greater than that of other diabetes drugs, was the safety of a drug like Parlodel in deciding tion, any corporate loss that results is likely to far higher than it would have been had the whether to keep prescribing it to his or her occur on the watch of future management.
defendant disclosed information relevant to patients. A clinician called as an expert wit- The executives who were present when prob- the safety of the product. Plaintiffs allege that ness who looks at all the available evidence lems first surfaced will probably long since they spent $1.4 billion purchasing Rezulin.
in reaching a decision about Parlodel's ability have cashed out. This time lag perhaps ex- The appellate court reviewing Desiano re- to cause strokes and heart attacks therefore plains why corporate officials may take a versed the district judge who had dismissed seems to fulfill the dictates of Kumho. The short-range view that is absolutely detrimen- the complaint for failing to state a claim. In role of clinicians in proving causation is, how- tal to a corporation's long-term objectives.
doing so, the appellate court noted that plain- ever, a question about which there is a good Perhaps the clearest example of this phenom- tiffs would not have to prove that Rezulin ac- deal of controversy.16 enon is the terrible but well-documented tually caused injuries. If plaintiffs ultimately Does Daubert promote deterrence? Mark- story of an AIDs-infected blood-clotting factor succeed in this law suit or obtain a favorable edly absent from this debate is a considera- for hemophiliacs that Cutter Biological never- settlement (keeping in mind that they seek tion of whether a stringent scientific standard theless kept selling overseas even though the damages in the neighborhood of one third of of causation furthers the deterrence objective product eventually killed its customers.20 $1.4 billion), it will be interesting to see of tort law and provides the public with ade- (Although the manufacturer of the blood whether a fraud claim not requiring proof of quate protection at a time when we are con- factor knew by October 1984 that heat treat- injury to users will induce pharmaceutical stantly being exposed to new unstudied sub- ment kills the AIDS virus, which it knew companies to divulge more information about stances and do not yet fully understand the could be transmitted through blood products, potential problems with their products than mechanisms that cause disease. It is, of course, it continued to ship the unheated product, they have provided under a regime in which true that the plaintiff bears the burden of which had become unmarketable in the liability requires proof of causation.
proof, but defendants are the ones creating United States, until the summer of 1985, the risk. However, even the regulatory admin- because it needed to use up its inventory of Daubert in Criminal Cases
istrative agencies, whose mandate is to deal Contrary to the very strict scrutiny some with risk rather than causation, are beginning Information about other pharmaceuticals federal courts apply when plaintiffs' experts to feel the impact of Daubert.17 that are the subject of toxic tort litigation fur- seek to testify, it should be noted that they I suggested a number of years ago, pre- ther supports the hypothesis that corporations seem to require considerably less with regard Enron, that stock options—now such a key may fail to acknowledge problems with their to prosecution experts in criminal cases. This component of executive compensation—create products. The Second Circuit Court's opinion is so even though in criminal proceedings, the disincentives for corporate behavior aimed at in Desiano v Warner-Lambert Co21 discusses prosecution must meet a much higher stan- preventing injury.18 Corporate executives do Rezulin, an anti-diabetes drug that was ap- dard of proof. The prosecution must prove its not wish to publicize adverse reports about proved by the FDA in January 1997 and was case beyond a reasonable doubt rather than their products for fear that stock prices will withdrawn from the U.S. market in March by a preponderance of the evidence. Never- tumble. News that a corporation is undertaking 2000 after the FDA concluded that Rezulin theless, although courts often exclude plain- additional research about an item already on usage "poses an unacceptable risk to pa- tiffs' experts in toxic tort cases, courts virtu- the market may be viewed as a sign of trouble.
tients."22 The defendant had apparently con- ally always permit a prosecution expert to Furthermore, as the results of any studies tinued to advertise its drug as safe, with "Side testify after a Daubert challenge. Much of the would be subject to discovery if litigation oc- Effects Comparable to Placebo," even after it expert proof in criminal cases consists of curred, it may be too dangerous to ask ques- became aware of fatalities in persons taking forensic identification testimony; it would tions to which no one knows the answer. For the drug and of reports linking severe liver seem far easier to test whether a given tech- corporations to take an ostrich-like approach is damage with Rezulin usage.23 nique can in fact match two tangible samples quite rational given our present system. A note It is, of course, impossible to know whether and to determine the frequency of such a from a corporate official was introduced in a the corporation's decision to promote the match than to decide whether a substance lawsuit claiming that the corporation deliber- drug despite its knowledge of adverse reac- can cause a particular disease. Clearly, how- ately avoided studying potential links between tions was in any way affected by the difficulty ever, except for a few cases excluding or lim- its product and a rare muscle disorder; the users of Rezulin would encounter in proving iting testimony about handwriting analysis,24 Supplement 1, 2005, Vol 95, No. S1 American Journal of Public Health Berger Peer Reviewed Public Health Matters S63 ⏐ PUBLIC HEALTH MATTERS ⏐ the courts are not applying Daubert strin- mony affects plaintiffs far more than defen- ment puts another potential obstacle in plain- gently in the criminal context. The paramount dants because plaintiffs may then not be able tiffs' path by perhaps driving out of the court- example is fingerprint evidence that has to meet their burden on proof. Furthermore, room good scientists who do not want to be never been validated. Although no one there is little point in plaintiffs going to the castigated as hired guns.
doubts that full sets of fingerprints can be expense of Daubert motions to exclude defen- Daubert's impact on transaction costs. The matched, the fingerprint found at a crime dant's experts until they know if their case RAND study also concluded that the rate of scene is often a partial, latent, contaminated will proceed. So if more experts are now exclusion of expert testimony has begun to print. How much of a print is needed for a being excluded, then Daubert has undoubt- drop and suggested that this may be due to match under these circumstances has not edly shifted the balance between plaintiffs better testimony being offered or to counsel been determined. Although a number of and defendants and made it more difficult for not pursuing cases in which Daubert stan- Daubert challenges have been made by de- plaintiffs to litigate successfully. Certainly, dards cannot be met. But there is an alterna- fense counsel, they have to date been uni- plaintiffs' counsel believe that this has hap- tive explanation. Plaintiffs may be warding off formly rejected. A federal judge who initially pened. When given a choice of whether to Daubert exclusions by offering more expert limited fingerprint experts to explaining simi- proceed in federal or state court, many plain- testimony than they used to and, accordingly, larities but barred them from expressing an tiffs now choose to bring suit in state court, are litigating only those cases in which the opinion about identity changed his mind.
where they think they will have a better potential award is large enough to cover these Compare United States v Llara Plaza, 179 chance of presenting their case to a jury.
additional expenses for experts. (The expense FSupp2d 492 (ED Pa 2002) with United Has the trilogy led to better expert proof? is more than just the experts' fees. For attor- States v Llara Plaza, 188 FSupp2d 549 (ED That, after all, was the rationale for the Su- neys paid on a contingent fee basis, as attor- Pa 2002). Both cases acknowledge the lack of preme Court's opinions. Nobody at this point neys in tort cases invariably are, Daubert un- research into the validity of matching finger- has the data to support such a conclusion, doubtedly forces them to spend much more prints; examiners have only been tested for because no one has as yet systematically com- time for which they cannot bill in preparing proficiency.25 When it comes to expert testi- pared proffered expert testimony that is ex- their experts. To explain to a court the mean- mony issues in criminal cases, the courts cluded with that which is admitted. The ing of statistical significance, for example, re- seem very conscious of the need to protect RAND Institute of Civil Justice has expressed quires understanding this concept and prepar- society against dangerous persons.
an interest in doing such work in the future, ing an expert to explain the standard of proof but whether this is a viable project remains to on which his or her opinion is based.) The Impact on Litigation in General be seen, given the thousands of pages of ex- Supreme Court's 2000 opinion in Weisgram v What conclusions can be drawn about the hibits that often accompany a Daubert motion Marley,29 which is sometimes referred to as general effects Daubert has had on civil litiga- and the disagreements among experts in dif- the Supreme Court's fourth case on expert tion? It needs to be noted that serious re- ferent disciplines.
testimony, helps force plaintiffs into this search on the effects of Daubert has just Impact on potential experts. Ironically, how- choice. In Weisgram, the trial court admitted begun and that it is difficult to do. Although ever, the trilogy may be making reputable sci- the plaintiffs' experts. Despite the abuse of cases in which judges exclude the plaintiffs' entists even more leery of participating in the discretion standard, the appellate court found experts and grant summary judgment can legal system. That a judge, who possibly has that the experts should have been excluded, generally be found because the trial court some incorrect or unsophisticated views about reversed, and entered judgment for defen- must write an opinion explaining its reasoning science, has the power to exclude the scientist dant. (In Kumho, the Court had once again if it is to avoid reversal for an "abuse of dis- as an expert witness and make some cutting endorsed the abuse of discretion standard of cretion," many cases in which plaintiffs win a remarks in print while doing so, may be review, which requires deference to a trial Daubert hearing undoubtedly settle without enough to convince some scientists that they court that admits expert testimony. However, being counted and disappear from sight, or do not wish to be involved with the legal sys- Justices O'Connor and Thomas joined Justice they go to trial and verdict without an opin- tem. And they may also for similar reasons Scalia in a brief concurring opinion to warn ion being written. The high visibility of deci- decline to undertake research related to litiga- that the abuse of discretion standard "is not sions that exclude plaintiffs' experts and grant tion. On the remand of Daubert, Judge Kozin- discretion to abandon the gatekeeping func- summary judgment may make the law appear ski of the Ninth Circuit added as a factor for tion" or "to perform the function inade- more settled than it actually is.
courts to consider in assessing reliability quately." It is not clear whether appellate Recent studies. Nevertheless, recent studies whether the expert's research was conducted courts will find an abuse of discretion more by the Federal Judicial Center26 and the expressly for the purpose of testifying and often when the trial court admits the plain- RAND Institute27 have concluded that judges suggested that unless science is conducted in- tiff's expert proof than when it excludes it.) are much more likely since Daubert to scruti- dependently of litigation, it is not likely to The plaintiffs argued that they were entitled nize expert testimony before trial and then to amount to "good science."28 But often the to a new trial at which they could produce limit or exclude expert testimony. As I stated need for research does not become apparent other experts. The Supreme Court did not re- at the outset, the exclusion of expert testi- until litigation begins. Judge Kozinski's assess- view the exclusion of the plaintiffs' experts by S64 Public Health Matters Peer Reviewed Berger American Journal of Public Health Supplement 1, 2005, Vol 95, No. S1 ⏐ PUBLIC HEALTH MATTERS ⏐ the appellate court; it dealt only with plain- expert testimony, in part because the Su- 15. Cohen NB. The gatekeeping role in civil litigation tiffs' claim to a new trial, which it rejected.
preme Court in another trilogy of opinions and the abdication of legal values in favor of scientificvalues. Seton Hall Law Rev. 2003;33:943,949–954.
"Since Daubert, moreover, parties relying on had made it easier for defendants to obtain expert evidence have had notice of the exact- summary judgment.32 Daubert works effec- 16. Kassirer JP, Cecil JS. Inconsistency in evidentiarystandards for medical testimony. JAMA. 2002;288: ing standards of reliability such evidence tively as another tool for terminating litigation must meet. . It is implausible to suggest, without a trial or jury.
17. Wagner WE. Importing Daubert to administrative post-Daubert, that parties will initially present agencies through the Information Quality Act. J Law less than their best expert evidence in the ex- Policy. 2004;12:589–617. See APHA Interim Policy
Statement LB03-1 (2003) on Threats to Public Health
pectation of a second chance should their first Science, Application of Supreme Court Decisions. Avail- try fail."30 (Of course, in contingent fee cases, The author is with Brooklyn Law School, Brooklyn, NY. able at: http://www.apha.org/legislative. Accessed June plaintiffs' lawyers who felt confident that the Requests for reprints should be sent to Margaret A. Berger, JD, Brooklyn Law School, 250 Joralemon Street, trial court would allow the plaintiffs' experts 18. Berger MA. Eliminating general causation: notes Room 911, Brooklyn, NY 11201 (e-mail: margaret.berger@ to testify might simply have been trying to towards a new theory of justice and toxic torts. Colum- bia Law Rev. 1997;97:2117–2152.
save money that would come out of their This article was accepted July 27, 2004. 19. Berenson A. Trial lawyers are now focusing on pockets if the lawsuit was unsuccessful; this is lawsuits against drug makers. New York Times. May 18, a gamble they can no longer afford to take.) Although hiring battalions of experts has This paper was supported, in part, by the Project on 20. Bogdanich W, Koli E. 2 Paths of Bayer Drugs in Scientific Knowledge and Public Policy.
improved plaintiffs' success rates and has pro- 80's: riskier one steered overseas. New York Times.
duced larger awards, clients with smaller May 22, 2003:A2.
claims may be unable to find representation.
21. Desiano v Warner Lambert Co, 326 F3d 339 (2d Daubert v Merrell Dow Pharmaceuticals, Inc, 509 It is expensive to proffer more experts at trial US 579 (1999).
than the trial court requires in the hope of 22. Desiano v Warner Lambert Co, 326 F3d 339, Mueller CB. Daubert asks the right questions: warding off a reversal on appeal. (It should now appellate courts should help find the right answer.
23. Desiano v Warner Lambert Co, 326 F3d 339, be noted that it is not only Daubert that has Seton Hall Law Rev. 2003;33:987–1023.
made expert proof more costly. Amendments Huber PW. Galileo's Revenge: Junk Science in the 24. See, for example, United States v Saelee, 162 to the Federal Rules of Civil Procedure that Courtroom. New York, NY: Basic Books; 1993.
FSupp2d 1097 (D Alaska 2001); United States v Fuji, became effective in 1993, the year Daubert Daubert v Merrell Dow Pharmaceuticals, Inc, 509 152 FSupp2d 939 (ND Ill 2000). Contra: United US 579, 585 (1993).
was decided, with the same objective of im- States v Prime, 220 FSupp2d 1203 (WD Wash 2002).
proving expert testimony, require testifying Frye v United States, 293 F 1013 (D.C. Cir. 1923).
25. See Zabell SL. Fingerprint evidence. J Law Policy.
2005;13:143–179.
experts to prepare written reports about their Barrett PM. Justices rule against business in evi- dence case. Wall Street Journal. July 29, 1993:A3.
opinions, after which they may be deposed.
26. Krafka C, Dunn M, Johnson MT, Cecil, JS,Miletich D. Judge and attorney experience, practices, All of this, as well as complying with Daubert Greenhouse L. Justices put judges in charge of de- ciding reliability of scientific testimony. Wall Street Jour- and concerns regarding expert testimony in federal and appearing at a Daubert hearing, means nal. July 29, 1993:A13.
civil trials. Psych Public Policy Law. 2002;8:309–332.
that the contingent fee lawyer must spend far Daubert v Merrell Dow Pharmaceuticals, Inc, 43 27 Dixon L, Gill B. Changes in the standards for ad- more time, which is not billable, on expert F3d 1311 (9th Cir 1995).
mitting expert evidence in federal civil cases since theDaubert decision. Santa Monica, Calif: RAND Institute testimony than before.) Carmichael v Samyang Tire, Inc, 131 F3d for Civil Justice; 2001.
One lawyer quoted in the New York Times 1433,1435 (11th Cir 1997), cert. granted sub nom.
28. Daubert v Merrell Dow Pharmaceuticals, Inc, explained, "I can no longer afford to spend Kumho Tire Co v Carmichael, 524 US 936 (1998), re-versed 526 US 137 (1999).
43 F3d 1311,1317 (9th Cir 1993).
$300,000 trying a case that is only worth 10. Kumho Tire Co v Carmichael, 526 U.S. 137,150 29. Weisgram v Marley, 528 US 440 (2000).
$500,000, and that's ridiculous."31 The num- 30. Weisgram v Marley, 528 US 440, 455 (2000).
ber of product liability cases filed in federal 11. Denbeaux MP, Risinger DM. Kumho Tire and ex- 31. Winter G. Jury awards soar as lawsuits decline on court has dropped precipitously. Daubert may pert reliability: how the question you ask gives the an- defective goods. New York Times. January 30, 2001:A1.
be stifling access to the courts at the same swer you get. Seton Hall Law Rev. 2003;34:15.
32. Matsushita Electric Industrial Co v Zenith Radio time it is fueling demands for tort reform by 12. Gatowski SI, Dobbin SA, Richardson JT, Ginsburg Corp, 475 US 574 (1986); Anderson v Liberty Lobby, escalating awards.
GP, Merino ML, Dahir V. Asking the gatekeepers: a na- Inc, 477 US 342 (1986); Celotex Corp v Catrett, 477 tional survey of judges on judging expert evidence in a The RAND report concluded, "Challenges US 317 (1986).
post-Daubert world. Law Hum Behav. 2001; to expert testimony increasingly resulted in summary judgment after Daubert," although 13. Peer Review Congress IV: A JAMA Theme Issue it also noted that "the increase may reflect (edited by Rennie D, Flanagan A). JAMA. 2002;287: broader trends in litigation practices that have little to do with Daubert." Certainly, the num- 14. Green MD, Freedman DM, Gordis L. Referenceguide on epidemiology. In: Reference Manual on Scien- ber of civil trials had been steadily decreasing tific Evidence. 2nd ed. Washington, DC: Federal Judicial even before the Supreme Court's opinions on Center; 2000.
Supplement 1, 2005, Vol 95, No. S1 American Journal of Public Health Berger Peer Reviewed Public Health Matters S65

Source: http://defendingscience.com/sites/default/files/upload/BergerDAUBERTWROUGHT.pdf